Riot, Ubisoft Collab on Research Initiative to Combat Toxic Comms Using AI

Riot Games and Ubisoft, two of the biggest names in the video game industry, have unveiled a new technological alliance dubbed the “Zero Harm in Comms” research project. 

To combat toxic and disruptive players, both firms will pool their resources to create a common database and labeling ecosystem that collects data from the game itself to use in training preemptive moderating systems.

With Riot’s extensive reach in the esports community and Ubisoft’s expertise in the gaming industry, the plan is to create a comprehensive database for their respective AI-based detection systems to analyze and identify all forms of disruptive activity.

The Partnership

Both companies are members of the Fair Play Alliance and share the same goal: to provide fair, safe, and inclusive environments in the expanse of online gaming. 

In a report by Dot Esports, Riot’s director of technical development, Wesley Kerr, acknowledged that disruptive conduct is not a challenge exclusive to games. 

According to him, every firm that offers an online social platform is attempting to handle this problematic space. 

This is why Riot is dedicated to collaborating with industry kingpins like Ubisoft that share their vision of promoting pleasant interactions and secure environments in virtual worlds.

Yves Jacquier, the head of Ubisoft’s La Forge division, said that Ubisoft has been working on tangible methods to guarantee safe and pleasant gaming encounters. Still, it believes the industry as a whole can better address the problem if everyone works together. 

“Through this technological partnership with Riot Games, we are exploring how to better prevent in-game toxicity as designers of these environments with a direct link to our communities.”

See Also: Riot Games Ends Partnership With Garena: Launches League of Legends, Teamfight Tactics Own SEA Server 

The Initiative

Ubisoft and Riot use chatlogs from their diverse online-focused games to train the AI at the heart of the Zero Harm in Comms initiative. 

This necessitates that their database has information on all the potential participants and behaviors that may be found in online frags and yeets.

The work focuses on AI and how to train it better to understand spoken language. 

Jacquier told Rock Paper Shotgun that traditional approaches give complete accuracy but are not scalable. The scalability of AI, on the other hand, comes at the cost of accuracy. 

Kerr added that traditional approaches have relied on utilizing AI to zero down on specific phrases. Nevertheless, this approach inevitably leaves out some disruptive behavior. 

He explained that with the developments in natural language processing and particularly some of the more recent big language models, they are seeing them be able to comprehend more context and subtlety in the language spoken rather than just searching for keywords.


Ubisoft and Riot want to present their findings to the video game industry by the beginning of the following year, regardless of the study’s results. 

This initiative ought to be simply the first step toward developing an even bigger cross-industry effort to assist the wider video gaming community for years to come if all is going according to plan.

Also Read: Ubisoft’s ‘Project U’ Leaks: Developer Teaser Talks About New Arena PvP Shooter, Not a Battle Royale

This article is owned by Tech Times

Written by Trisha Kae Andrada

ⓒ 2022 All rights reserved. Do not reproduce without permission.

Leave a Reply

Your email address will not be published. Required fields are marked *